14:56:53 #startmeeting Copper Weekly Meeting 14:56:53 Meeting started Wed Aug 5 14:56:53 2015 UTC. The chair is bryan_att. Information about MeetBot at http://wiki.debian.org/MeetBot. 14:56:53 Useful Commands: #action #agreed #help #info #idea #link #topic. 14:56:53 The meeting name has been set to 'copper_weekly_meeting' 14:57:06 #info Bryan Sullivan 14:58:15 #info No prepared agenda for today, open discussion if any has any topics following the OPNFV summit 15:00:01 #info I'm focused on gettine a testbed off the ground (JIra: 2) 15:00:05 #link https://jira.opnfv.org/browse/COPPER-2 15:01:34 #info this is the first key prerequisite to having a demo capability for the Nov summit and for Congress as part of the B release 15:02:11 I'll wait a few more minutes online to see if anyone else joins; ping me if I don't seem to be here 15:03:29 #info I am new to this project. I see lot of relevance between what we are trying to do with VNF Manager and COPPER requirements. Am i reading this right? 15:05:10 #info No sure, what specific relation do you see? Clearly VNFM has a role (or there is a VNF management related role, more generally) for setting up resources per the needs of a NF 15:05:53 #info as part of the upper MANO stack functions; but what specifically do you have in mind? 15:06:06 Policy 15:06:28 As the project description says - consolidate policy functionality from Congress and GBP 15:06:50 MANO/Tacker has policy functionality as well - more like enforcement 15:07:01 #info there are several aspects to policy; the project broadly covers them, but will focus on the ones that are unique to it 15:08:09 #info for example; knowing what is needed for a NF per resources, and what is needed for a NS as a chain of NFs, and then fulfilling those needs thru VIM APIs - that is "configuration intent" 15:08:22 #info are we going to have GoToMeeting for this call? 15:08:36 no GTM - only IRC 15:08:43 ok 15:09:33 #info whereas making sure that things don't get done that violate some generic (VNF-independent) policy; that is "config enforcement" 15:10:27 #info "config intent" definition and fulfillment is a broader topic that I am trying not to duplicate in Copper, though it clearly is a dependency for the system overall 15:11:20 #info e.g. we need a mechanism for expressing intent "what" and "how" levels so we can also ensure that we can express what *should not* happen (config violations) 15:11:41 #info is that clearer re the scope of Copper and VNFM? 15:12:04 #info Do you see GBP as a place for 'config intent'? 15:12:58 #info Yes, GBP is one mechanism that can fulfill grouping NFs for configuration into services 15:13:43 #info but to avoid stretching the term; GBP really relates to "how" rather than "what"; when you hear others speak of intent they mean more the "what" 15:14:18 #info i.e. "intent" means to them an implementaton-approach-independent express of what is wanted, rather than how it's achieved 15:14:26 #info So you are saying intent 'what' and 'how' is what is covered in Copper? 15:15:13 #info Both what and how are covered in Copper, but more focused on ensuring config policy violations are detectable and can be dealt with 15:15:55 #info sounds like traditional PM in FCAPS but on policy - Policy Performance 15:16:12 #info Re how things are supposed to be built and how that relates to what was originally expressed as a need (what), that's a broader topic that Copper will depend upon but not address specifically 15:17:13 #info Unless its found that these are not being addressed by other projects... 15:18:26 #info And then we will take config intent and fulfillment in Copper, e.g. as i have proposed for the summit demo on "A day in the life of a VNF" whch will touch on and demo VNFM and NFVO functions 15:20:05 #infoo but for now, config enforcement is the prime focus; thus getting Congress into the OPNFV build is a first priority; ODL GBP/SFC is next; etc (all the dependent functions that can be used to define how things are *supposed* to work can also be useful for defining what *should not* happen) 15:20:18 #info but for now, config enforcement is the prime focus; thus getting Congress into the OPNFV build is a first priority; ODL GBP/SFC is next; etc (all the dependent functions that can be used to define how things are *supposed* to work can also be useful for defining what *should not* happen) 15:20:48 Any other questions? Otherwise I can give you back the time. 15:21:09 #info Is any one working on this already towards the Nov demo? 15:21:35 #info I am working to get a lab setup - that's task #1 15:22:05 #info I can contribute as well. Please point me to the right resources to get started. 15:22:15 #info It would be great to have that being done in multiple places - i.e. labs - some competition to getting Congress installed 15:22:33 #info Do you have access to a Pharos lab? 15:22:55 I am trying to get from Dell. Looks like they are backed up until Sept. 15:24:17 #info That's the first step - you need a lab that you can then install the latest stable Kilo and Lithium on, then enhance with Congress and SFC/GBP (enventually) 15:25:33 #info will try to setup one in my company. Thanks for the info. Let me know if there are other ways I can contribute as well. Talk to you next Week. 15:25:44 #info As I develop any extra docs needed to get Congress installed (beyond their docs) i will post them on the wiki 15:25:55 Thanks, talk to you later 15:25:59 #endmeeting